|Home | About | Journals | Submit | Contact Us | Français|
We would like to acknowledge the previous contributor of this review: Peter McCulloch.
Stomach cancer is usually an adenocarcinoma arising in the stomach, and includes tumours arising at or just below the gastro-oesophageal junction (type II and III junctional tumours). The annual incidence varies among countries and by sex, with about 80 cases a year per 100,000 in Japanese men, 30/100,000 in Japanese women, 18/100,000 in British men, and 10/100,000 in British women.
We conducted a systematic review and aimed to answer the following clinical questions: What are the effects of radical versus conservative surgical resection? What are the effects of adjuvant chemotherapy? We searched: Medline, Embase, The Cochrane Library, and other important databases up to August 2007 (BMJ Clinical Evidence reviews are updated periodically; please check our website for the most up-to-date version of this review). We included harms alerts from relevant organisations such as the US Food and Drug Administration (FDA) and the UK Medicines and Healthcare products Regulatory Agency (MHRA).
We found 18 systematic reviews, RCTs, or observational studies that met our inclusion criteria. We performed a GRADE evaluation of the quality of evidence for interventions.
In this systematic review we present information relating to the effectiveness and safety of the following interventions: adjuvant chemoradiotherapy, adjuvant chemotherapy, lymphadenectomy (radical, conservative), removal of adjacent organs, and subtotal gastrectomy for resectable distal tumours.
Stomach cancer is usually an adenocarcinoma arising in the stomach and includes tumours arising at or just below the gastro-oesophageal junction (type II and III junctional tumours). Only non-metastatic stomach cancers are considered in this review.
With regard to surgical resection, subtotal gastrectomy seems as effective as total gastrectomy.
Removal of adjacent organs (spleen and distal pancreas) is associated with increased morbidity and mortality compared with gastrectomy alone.
We found insufficient evidence to judge the effectiveness of radical lymphadenectomy compared with conservative lymphadenectomy.
Adjuvant chemoradiotherapy seems to improve survival compared with surgery alone in people with resectable stomach adenocarcinoma.
Adjuvant chemotherapy might also be effective compared with surgery alone, although the evidence is inconsistent.
Stomach cancer is usually an adenocarcinoma arising in the stomach, and includes tumours arising at or just below the gastro-oesophageal junction (type II and III junctional tumours). Tumours are staged according to degree of invasion and spread (see table 1 ). Only non-metastatic stomach cancers are considered in this review.
The incidence of stomach cancer varies among countries and by sex (incidence per 100,000 population a year in Japanese men is about 80, Japanese women 30, British men 18, British women 10, white American men 11, and white American women 7). Incidence has declined dramatically in North America, Australia, and New Zealand since 1930, but the decline in Europe has been slower. In the USA, stomach cancer remains relatively common among particular ethnic groups, especially Japanese-Americans, and some Hispanic groups. The incidence of cancer of the proximal stomach and gastro-oesophageal junction is rising rapidly in many European populations and in North America. The reasons for this are poorly understood.
Distal stomach cancer is strongly associated with lifelong infection with Helicobacter pylori and poor dietary intake of antioxidant vitamins (A, C, and E). In Western Europe and North America, distal stomach cancer is associated with relative socioeconomic deprivation. Proximal stomach cancer is strongly associated with smoking (OR about 4), and is probably associated with gastro-oesophageal reflux, obesity, high fat intake, and medium to high socioeconomic status.
Invasive stomach cancer (stages T2-T4) is fatal without surgery. Mean survival without treatment is less than 6 months from diagnosis. Intramucosal or submucosal cancer (stage T1) may progress slowly to invasive cancer over several years. In the USA, over 50% of people recently diagnosed with stomach cancer have regional lymph node metastasis or involvement of adjacent organs. The prognosis after macroscopically and microscopically complete resection (R0) is related strongly to disease stage, particularly penetration of the serosa (stage T3) and lymph node involvement. Five-year survival rates range from over 90% in intramucosal cancer to about 20% in people with stage T3N2 disease (see table 1 ). In Japan, the 5-year survival rate for people with advanced disease is reported to be about 50%, but the explanation for the difference remains unclear. Comparisons between Japanese and Western practice are confounded by factors such as age, fitness, and disease stage, as well as by tumour location, because many Western series include gastro-oesophageal junction adenocarcinoma, which is associated with a much lower survival rate after surgery.
To prevent progression; extend survival; and relieve symptoms, with minimal adverse effects.
Survival; quality of life; adverse effects of treatment.
BMJ Clinical Evidence search and appraisal August 2007. Additional searches were carried out using these websites: NHS Centre for Reviews and Dissemination (CRD) — for Database of Abstracts of Reviews of Effects (DARE) and Health Technology Assessment (HTA), Turning Research into Practice (TRIP), and NICE clinical guidelines. The author also performed hand searches of conference proceedings and consultations with experts to identify relevant studies in August 2004. Abstracts of the studies retrieved were assessed independently by two information specialists using predetermined criteria to identify relevant studies. Study design criteria for inclusion in this review were: published systematic reviews and RCTs in any language, at least single blinded, and containing more than 20 people, of whom more than 80% were followed up. There was no minimum length of follow-up required to include studies. We excluded all studies described as "open", "open label", or not blinded unless blinding was impossible. In addition, we use a regular surveillance protocol to capture harms alerts from organisations such as the FDA and the UK Medicines and Healthcare products Regulatory Agency (MHRA), which are added to the review as required. In many instances, we have separated trials and results from different geographical areas, because of differences in baseline risk and demographics and possible differences in responses to treatments. However, the meanings of terms to describe such populations, such as "Western" and "Asian" were not clearly defined in many identified studies. We have performed a GRADE evaluation of the quality of evidence for interventions included in this review (see table ).
The information contained in this publication is intended for medical professionals. Categories presented in Clinical Evidence indicate a judgement about the strength of the evidence available to our contributors prior to publication and the relevant importance of benefit and harms. We rely on our contributors to confirm the accuracy of the information presented and to adhere to describe accepted practices. Readers should be aware that professionals in the field may have different opinions. Because of this and regular advances in medical research we strongly recommend that readers' independently verify specified treatments and drugs including manufacturers' guidance. Also, the categories do not indicate whether a particular treatment is generally appropriate or whether it is suitable for a particular individual. Ultimately it is the readers' responsibility to make their own professional judgements, so to appropriately advise and treat their patients.To the fullest extent permitted by law, BMJ Publishing Group Limited and its editors are not responsible for any losses, injury or damage caused to any person or property (including under contract, by negligence, products liability or otherwise) whether they be direct or indirect, special, incidental or consequential, resulting from the application of the information in this publication.
MORTALITY Subtotal compared with total gastrectomy: We don't know whether subtotal gastrectomy is more effective at increasing survival at 5 years in people with primary tumours in the distal stomach ( moderate-quality evidence ).
We found no systematic review but found two RCTs (787 people, 4 publications) comparing total versus subtotal gastrectomy. Neither RCT used blinded allocation. The first RCT (648 people aged less than 76 years with a resectable tumour and a macroscopic proximal margin of more than 6 cm) compared total versus subtotal gastrectomy. All people involved in the RCT had a regional lymphadenectomy (D2; see table 2 ) The RCT found no significant difference between total and subtotal gastrectomy in the incidence of microscopic resection margin involvement (15/315 [5%] with subtotal gastrectomy v 6/303 [2%] with total gastrectomy; ARI +2.8%, 95% CI –0.1% to +9.6%) or in 5-year survival (Kaplan–Meier 5-year survival estimates: 65% with subtotal gastrectomy v 62% with total gastrectomy; HR 0.89, 95% CI 0.68 to 1.17). Multivariate analysis found that, after adjustment for covariates, the type of stomach surgery had no significant effect on 5-year survival (HR 1.01, 95% CI 0.76 to 1.33). The second RCT (169 people with potentially curable distal stomach cancer) compared total versus subtotal gastrectomy, and found no significant difference in 5-year survival (48% in each group; CI not reported).
We found no RCTs that examined quality of life with total compared with subtotal gastrectomy in people with resectable distal tumours.
Morbidity included intra-abdominal sepsis, chest infections, wound sepsis, and fistulae. The first RCT found no significant difference between subtotal and total gastrectomy in postoperative morbidity (29/320 [9%] with subtotal gastrectomy v 40/304 [13%] with total gastrectomy; RR 0.67, 95% CI 0.44 to 1.08), although length of hospital stay was significantly longer in the total gastrectomy group (13.8 days with subtotal gastrectomy v 15.4 days with total gastrectomy, P less than 0.001). The second RCT also found no significant difference in postoperative morbidity (32/93 [34%] with subtotal gastrectomy v 25/76 [32%] with total gastrectomy; RR 1.05, 95% CI 0.68 to 1.60).
The first RCT found no significant difference in postoperative mortality (4/320 [1%] with subtotal gastrectomy v 7/304 [2%] with total gastrectomy; RR 0.54, 95% CI 0.16 to 1.84). The second RCT also found no significant difference in postoperative mortality (3/93 [3%] with subtotal gastrectomy v 1/76 [1%] with total gastrectomy; RR 2.45, 95% CI 0.26 to 23.01).
Infiltration of the proximal resection margin by microscopic tumour deposits is perceived as a problem in people with poorly differentiated “diffuse” cancer of the distal stomach having distal subtotal gastrectomy. Some surgeons have therefore recommended total gastrectomy “de principe” for these tumours. The two RCTs found similar survival after total and subtotal gastrectomy in people with primary tumours in the distal stomach. Both RCTs recruited otherwise-fit people, which may explain the low postoperative mortality.
No new evidence
MORTALITY Radical compared with conservative lymphadenectomy: We don't know whether radical lymphadenectomy is more effective at increasing survival at 5 or 10 years ( low-quality evidence ).
We found one systematic review (search date 2003, 2 RCTs, 1111 people), which found no significant difference in 5-year survival between radical and conservative lymphadenectomy (AR: 237/580 [41%] with conservative lymphadenectomy v 226/531 [43%] with radical lymphadenectomy; RR 0.95, 95% CI 0.83 to 1.09). Ten-year follow-up of one of the RCTs included in the review found similar results (AR for survival: 30% with conservative lymphadenectomy v 35% with radical lymphadenectomy; P = 0.53).
The systematic review found that radical lymphadenectomy significantly increased postoperative mortality compared with conservative lymphadenectomy (AR: 28/572 [5%] with conservative lymphadenectomy v 58/523 [11%] with radical lymphadenectomy; RR 2.23, 95% CI 1.45 to 3.46). The excess postoperative mortality may have been due to pancreatic and splenic removal associated with radical lymphadenectomy. The 10-year follow-up of one of the RCTs included in the review did not report on harms; they were described in the original trial publication.
The RCTs were conducted by surgeons with limited prior experience and training in radical lymphadenectomy (D2 resection), and results may have been affected by both learning-curve effects, and failure to apply the assigned treatment (contamination and non-compliance). Subgroup analysis from one of the RCTs included in the review found a possible advantage for D2 resection in people with stage II and IIIA disease (corresponding to T1N2M0, T2N1M0, T3N0M0, T2N2M0, T3N1M0, and T4N0M0), particularly in those people who did not have additional organ removal. One large prospective cohort study (1654 people with gastric cancer) found no benefit from D2 resection (defined within this study as more than 25 lymph nodes removed; 300 people) compared with D1 resection (up to 25 nodes removed; 1096 people) in the entire cohort of people with gastric cancer after 10 years' follow-up. Subgroup analysis found that there may be a beneficial effect of D2 compared with D1 resection in the subgroup of people with stage II tumours (230 people; RR of long-term survival 1.8, 95% CI 1.3 to 2.7). Cohort studies comparing radical versus conservative lymphadenectomy are affected by numerous biases: selection bias (where surgeons reserve D2 surgery for younger or fitter people, or where recent D2 operations are compared with historical D1 controls); differing definitions of “limited” and “extended”; and stage migration bias. These biases make the interpretation of observational data difficult.
No new evidence
MORTALITY Total gastrectomy plus splenectomy compared with total gastrectomy: Total gastrectomy plus splenectomy and total gastrectomy alone are equally effective at increasing survival at 5 years ( moderate-quality evidence ). NOTE There is consensus that removal of adjacent organs is justified to ensure complete tumour removal, or when required because of trauma during surgery.
We found no systematic review but found two RCTs. The first RCT (187 people aged 25–80 years) found no significant difference in 5-year survival rates for total gastrectomy plus splenectomy compared with total gastrectomy alone (42% with total gastrectomy plus splenectomy v 36% with total gastrectomy alone; P greater than 0.5; absolute numbers not reported). The second RCT (216 people, aged 24–78 years) compared total gastrectomy plus splenectomy versus total gastrectomy alone. It also found no significant difference in 5-year survival between the two groups (54.8% with total gastrectomy plus splenectomy v 48.8% with total gastrectomy alone; P = 0.503; absolute numbers not reported).
The first RCT (187 people) found that total gastrectomy plus splenectomy significantly increased the number of postoperative infections compared with total gastrectomy alone (fever above 38 °C, P less than 0.04; pulmonary infections, P less than 0.008; subphrenic abscesses, P less than 0.05), but found no significant difference in postoperative mortality (4/90 [4%] with total gastrectomy plus splenectomy v 3/97 [3%] with total gastrectomy alone; RR 1.40, 95% CI 0.33 to 6.24). The second RCT showed no significant difference in postoperative complications (15.4% with total gastrectomy plus splenectomy v 8.7% with total gastrectomy alone; P = 0.142). Retrospective analyses of RCTs and cohort studies, in which removal of the spleen and distal pancreas had been carried out routinely during radical total gastrectomy (D2) (see table 2 ), and at the surgeon's discretion during non-radical total gastrectomy (D1), found that removal of the spleen or distal pancreas was associated with increased perioperative mortality (OR about 2), and found no evidence of improved long-term survival.
Some advocates of radical surgery have suggested routine removal of the spleen and distal pancreas to ensure complete regional lymph node dissection during total gastrectomy. Current consensus is that removal of adjacent organs is justified only when necessary to ensure complete tumour removal, or when required because of trauma during surgery. One RCT has begun in Japan to evaluate the role of splenectomy in total gastrectomy for proximal gastric cancer.
No new evidence
MORTALITY Compared with surgery: Adjuvant chemoradiotherapy seems more effective at increasing survival in people with resectable stomach adenocarcinoma ( moderate-quality evidence ).
We found no systematic review, but found one RCT comparing adjuvant chemoradiotherapy versus surgery alone for resectable stomach adenocarcinoma. It found that adjuvant chemoradiotherapy significantly increased survival compared with surgery alone (556 people; median survival: 36 months with surgery plus chemoradiotherapy v 27 months with surgery alone; HR 1.35, 95% CI 1.09 to 1.66).
The RCT reported three treatment-related deaths in the adjuvant chemoradiotherapy group (3/273 [1%]). Grade 3 toxic effects occurred in 41% of people receiving chemoradiotherapy and grade 4 toxic effects in 32%. The RCT gave no information on harms in the surgery alone group.
Concerns have been raised about the lack of standardisation of the surgical resection technique. As surgery was not standardised, it is possible that resection was not thorough enough in some people, leaving them with residual disease — so that they may have benefited from chemoradiotherapy compared with no treatment. If the surgery had been thorough (i.e. standardised), adjuvant chemoradiotherapy may not have been found to be beneficial compared with no treatment.
No new evidence
MORTALITY Compared with surgery: Adjuvant chemotherapy seems more effective at improving overall survival in people with gastric adenocarcinoma, but may be poorly tolerated ( moderate-quality evidence ).
We found two systematic reviews and six subsequent RCTs, comparing adjuvant chemotherapy versus surgery alone. The first review (search date 2000, 20 RCTs, 3658 people; see comment below) found that adjuvant chemotherapy significantly reduced mortality compared with surgery alone (HR 0.82, 95% CI 0.75 to 0.89). The second review (search date 2005, 4 RCTs, 1330 people) was confined to trials performed in Japan (none of which were included in the first review), and also found that adjuvant chemotherapy significantly reduced mortality compared with surgery alone (HR 0.73, 95% CI 0.60 to 0.89, no absolute data reported). The first subsequent RCT (137 people with gastric adenocarcinoma and positive lymph nodes) compared adjuvant chemotherapy versus surgery alone, and found that adjuvant chemotherapy significantly increased median survival time (31 months [range 7 months to greater than 60 months] with adjuvant chemotherapy v 18 months [range 2 months to greater than 60 months] with surgery alone; P less than 0.01; HR for death 1.96, 95% CI 1.32 to 2.92). The second subsequent RCT (274 people with gastric adenocarcinoma T3, T4 or N1, N2) found no significant difference in 5-year survival between adjuvant chemotherapy and surgery alone (overall survival: 52% with adjuvant chemotherapy v 48% with surgery alone; HR 0.93, 95% CI 0.65 to 1.34). The third subsequent RCT (139 people) compared three groups: surgery plus hyperthermic intraperitoneal chemotherapy; surgery plus normothermic intraperitoneal chemotherapy; and surgery alone. It found that surgery plus hyperthermic chemotherapy (mitomycin C plus cisplatin at 42 °C) significantly increased overall 5-year survival rates compared with surgery alone (P = 0.01), and that surgery plus hyperthermic chemotherapy also significantly increased overall 5-year survival rates compared with surgery plus normothermic chemotherapy (mitomycin C plus cisplatin at 37 °C) (P = 0.05; absolute results presented graphically). Subgroup analysis found that these results were consistent for people with advanced disease (T3 or node positive), but, in people with less-advanced disease (T2 or node negative), subgroup analysis found no significant difference in overall 5-year survival rates. The fourth subsequent RCT (205 people with serosa- or lymph node-positive gastric cancer) also found no significant difference in 5-year survival between adjuvant chemotherapy and surgery alone (39% with adjuvant chemotherapy v 39% with surgery alone). However, owing to treatment-related toxicity, 54% of people did not complete the course of chemotherapy, and 32% completed chemotherapy at a decreased dose. The fifth subsequent RCT of gastro-oesophageal cancer (including cancers of the stomach [372 people], gastro-oesophageal junction [58 people], and lower third of the oesophagus [73 people]) compared perioperative chemotherapy plus surgery versus surgery alone. It found significantly improved rates of progression-free and overall survival for perioperative chemotherapy compared with surgery alone at a median follow-up of 4 years (progression-free survival: HR 0.66, 95% CI 0.53 to 0.81, P less than 0.001; overall survival: 149/250 [60%] deaths with perioperative chemotherapy v 170/253 [67%] deaths with surgery alone; HR for death 0.75, 95% CI 0.60 to 0.93, P = 0.009). The sixth subsequent RCT (40 people with locally advanced gastric tumours (T3 and T4) without distant metastases) compared surgery plus intra-arterial chemotherapy versus surgery alone. The RCT found no significant difference between groups in mean survival (50 months 95% CI 42 to 81 months with surgery plus intra-arterial chemotherapy v 62 months, 95% CI 34 to 66 months with surgery alone, P = 0.9).
Two RCTs included in the review reported toxicity (mainly nausea and vomiting) in 53% of people. Serious toxicity was usually because of cardiac or cumulative haematological problems; treatment-related mortality was 1–2%. One subsequent RCTs reported no significant difference in postoperative complications with adjuvant chemotherapy. One subsequent RCT reported 4/93 (4%) chemotherapy-related deaths. Most people in the RCT stopped or reduced chemotherapy, because of adverse effects. The RCT of gastro-oesophageal cancers reported no clinically significant increase in the incidence of grade 3 or grade 4 toxic effects.The sixth subsequent RCT reported minor toxicity in three people receiving surgery plus intra-arterial chemotherapy (no further data reported).
Two RCTs included in the systematic review may have been duplicate versions of the same RCT. Preoperative superselective intra-arterial chemotherapy may not be available outside of specialist centres.
Adjuvant chemotherapy One systematic review and one subsequent RCT added comparing adjuvant chemotherapy versus surgery alone. The review found that adjuvant chemotherapy reduced mortality compared with surgery, but the subsequent RCT found no significant difference in mean survival rates between adjuvant chemotherapy and surgery alone. Categorisation unchanged (Likely to be beneficial).